NW/. Med. Bid. Vol. 18, No. 6, pp. 641-645, Im. J. Rod&. Appl. Instrum. Part B
0883s2897/91 $3.00+ 0.00 PergamonPressplc
1991
Printed in Great Britain.
Radionuclide Assessment of Blood-Brain Barrier Disruption Performed for Chemotherapy of High Grade Malignant Brain Gliomas AMOLAK
SINGH*,
BARBARA D. BELSHE and MARY K. GUMERLOCK
Departments of Radiology (Nuclear Medicine) and Neurosurgery, University of Missouri Hospital and Clinics, Columbia, MO 65212, U.S.A.
Compared to the conventional mode of chemotherapy of malignant brain gliomas following surgery and radiation therapy, chemotherapy after transient disruption of the blood-brain barrier coupled with intraarterial administration of methotrexate improved median survival from 12-14 to 22 months in our experience. Technetium-99m-DTPA brain scintigraphy played a unique and important role in the documentation of optimum blood-brain barrier disruption. Patients with excellent clinical outcome had significantly (P < 0.0005) better blood-brain barrier disruption than patients with poor outcome. The results indicate that the clinical outcome is related to the degree of blood-brain barrier disruption.
Introduction High grade malignant brain gliomas after initial surgical removal and radiation therapy are generally treated with chemotherapeutic agents which are administered orally or intravenously. Such therapeutic regimens have not significantly altered survival in these patients. At the University of Missouri Hospital and Clinics, the high grade malignant brain gliomas after initial surgery and irradiation are treated with chemotherapeutic drugs which are administered following transient disruption of the of the blood-brain barrier (BBB) using hyperosmolar intraarterial mannitol. By opening the blood-brain barrier at the time of chemotherapy the drug delivery to the tumor is enhanced not only by increasing concentration of the drug within the tumor mass but also by eliminating the steep concentration gradient which exists between tumor and surrounding brain tissue. As a result of this anti “SUMP” effect, the drug-tumor contact is prolonged. Moreover, to maximize tumor bombardment with chemotherapy, methotrexate is injected intraarterially directly into the blood supply of the tumor. The disruption of the blood-brain barrier is evaluated using 99mTc-diethylenetriaminepentaacetic acid (DTPA) brain scintigraphy to monitor this
*All correspondence should be addressed to: Amolak Singh, Division of Nuclear Medicine, One Hospital Drive, 2N-19 Columbia, MO 65212, U.S.A.
disruption tumor.
for optimum
drug chemotherapy
to the
Materials and Methods The study population included 37 patients with histologically proven high grade malignant brain gliomas. There were 15 female and 22 males, with a mean age of 43 years, ranging from 12 to 71 years. The functional condition of the patient was established by using Karnofsky’s performance (KP) score (Schag et al., 1984). A total of 246 chemotherapeutic treatments were instituted in 37 patients. (A) Technique for (BBBD)
blood-brain
barrier
disruption
BBBD was achieved using a hyperosmolar (25%) solution of mannitol injected into the internal carotid and/or vertebral arteries depending upon tumor blood supply. The rate of mannitol infusion varied from 5 to 12 mL/s for a 30 s period. Methotrexate (1-5 g) was infused intraarterially into the tumor blood supply either through the internal carotid artery or vertebral artery, or both in some cases, immediately following BBBD with hyperosmolar mannitol. Cyclophosphamide (20 mg/kg) was given intravenously 10 min prior to BBBD with mannitol to allow sufficient time for hepatic circulation and drug activation, Procarbazine (100 mg) was given orally daily for 14 days starting the day after disruption. 641
642
AMOLAK SINGH et al.
Leucovorin, a folic acid derivative, was used to counteract toxic effects of methotrexate. The patients underwent tumor BBBD-chemotherapy every 4 + weeks guided by clinical response, brain computed tomography (CT) and other laboratory studies done prior to each planned BBBD chemotherapy. (B) Technetium-9%
DTPA scintigraphy
All patients underwent a baseline brain scan and BBBD-brain scans during therapy. BBBD-brain scintigrams were obtained using 20mCi of 99mTcDTPA administered intravenously immediately following mannitol infusion. Anterior posterior, vertex and both lateral brain images were obtained 2-3 h post-injection of the radiopharmaceutical using a gamma camera (Siemens 7500) interfaced with a nuclear medicine computer (ADAC-3300). Occasionally, SPECT images were also generated in addition to routine planar images. Using the 99mTc-DTPA brain scintigram, the degree of BBBD was routinely evaluated using a grading system based on the presence or absence of increased brain uptake in the region outside the tumor margins. Absence of disruption was graded as 0, minimal uptake as I+, moderate uptake as 2 +, and 3 + as intense uptake equal to or greater than venous sinuses. The BBBD index, defined as the ratio of count density from the brain tissue in the vicinity but outside tumor margin to contralateral normal brain tissue, was determined. The BBBD index in patients with excellent outcome was compared with BBBD index in patients with poor outcome using Wilcoxon rank sum test. A P value co.05 was required for significant difference. All data were expressed as meanf 1 SD.
Results Using CT scan as a diagnostic test for evaluation of tumor size before therapy and subsequently during follow-up examinations, six (16%) patients were found to have excellent tumor response and were found alive 37-78 months following diagnosis. Twenty-four (65%) patients had partial or temporary remission. Seven (19%) had poor outcome. Calculated median survival time was 22 months. One hundred and forty-two (65%) out of 232 studies were judged to have adequate (2 + or 3 +) disruption, 54 (23%) had minimal (1 +) disruption and 27 (12%) had no disruption. BBBD was usually adequate with carotid artery injection of mannitol (72%) versus
vertebral artery injection in which case only 20% of the patients revealed adequate disruption. As shown in Table 1, patients with excellent response (Patient No. = 6, studies = 36) revealed a mean BBBD index of 1.94 f 0.25 compared to patients with poor outcome (Patient No. = 7, studies = 25) who exhibited a mean BBBD index of 1.42 + 0.25. The difference was highly significant (P c 0.0005). Representative cases (1) A 30-year-old female presented with a left frontal lobe high grade malignant glioma. A pretherapy CT scan revealed postoperative changes and residual tumor in the left frontal lobe [Fig. l(A)]. A BBBD scintigram demonstrated excellent (3 i-) BBBD at the time of chemotherapy [Fig. l(B)]. The patient underwent eleven such mannitol-induced BBBD and treatments. A follow-up CT scan showed complete resolution of the tumor [Fig. l(C)]. This patient was alive and well 4 years after the first treatment. (2) A 50-year-old female was included in the protocol for therapy of a large parietal lobe high glioma [Fig. 2(A)]. A BBBD brain scintigram demonstrated no significant BBBD outside the tumor margins [Fig. 2(B)]. The patient underwent three such mannitolinduced BBBD and chemotherapies without any response. A follow-up CT scan [Fig. 2(C)] exhibited further enlargement of the tumor. This patient died within a year following inclusion in the protocol.
Discussion Conventional mode of chemotherapy after surgical removal and irradiation therapy of high grade malig nant brain gliomas has not been successful in altering survival in these patients. Our results with BBBD chemotherapy indicate significant improvement in patient survival with increase in median survival time from 12-14 to 22 months. This compares favorably with 16.5 months reported by Neuwelt et al. (1986). Technetium-99m-DTPA brain scintigraphy plays an important role in the assessment of the degree of BBBD and documentation of optimum drug delivery to the area of interest (tumor) and the surrounding brain tissue. Moreover, scintigraphic studies were helpful for guiding mannitol infusion rates required for optimum BBBD. Adequate (2+ to 3+) BBBD occurred in 65% of our studies. No disruption was noted in 12%; in the remaining
Table I. Relationship between clinical outcome and degree of BBBD* Clinical outcome
BBBD Index (mean + SD)
Excellent response (Number of patients: 6; number of studies: 36)
1.94 f 0.44
Poor response (number of oatients: 7: number of studies: 25)
I .42 + 0.25
Pt
and of questionable significance. Internal carotid artery injections of mannitol produced adequate disruption in 72% of studies as compared to the vertebral artery injection in which case only 20% of the studies revealed adequate disruption. These results suggest that improved delivery of chemotherapeutic drugs to the region of interest as a result of BBBD is achieved in approximately three fourths (3/4) of treatments following mannitol injections through internal carotid artery, and in only one fifth (l/5) of treatments following vertebral artery injections of mannitol. However, all cases benefit from intraarterial administration of chemotherapeutic drugs. The degree of BBBD and clinical outcome appear related since patients with excellent outcome have better BBBD than those with poor outcome (P < 0.0005). This finding has not been previously reported. This application of radionuclide brain scan using 99mT~labeled chelates represents a unique application of conventional nondiffusible brain agents. Newer diffusible brain agents such as 99mTc-hexamethylpropyleneamine oxide (HMPAO) and ‘XI-labeled amphetamine derivatives are not suitable for this application. Moreover. the radionuclide brain scan is superior to CT in the assessment of BBBD because of higher safety. CT scans performed for the assessment
of BBBD are associated with fairly high (17%) incidence of contrast induced seizures in patients undergoing BBBD and chemotherapeutic treatments (Neuwelt et al., 1983). In conclusion, BBBD chemotherapy of malignant brain gliomas significantly improves survival compared to the conventional mode of chemotherapy. Technetium-99m DTPA (or other similar agents such as WmTc-glucoheptonate) brain scintigraphy is the technique of choice for the assessment of blood-brain barrier disruption, and for documentation of optimum drug delivery to the region of interest. The degree of BBBD as measured by brain scintigraphy appears related to survival and clinical outcome in these patients with high malignant brain glioma.
References Neuwelt E. A., Specht D. H., Howieson J. et al. (1983) Osmotic blood-brain-barrier modification: clinical documentation by enhanced CT scanning and/or radionuclide brain scanning. Am. J. Radiol. 4, 907-913. Neuwelt E. A., Howieson J., Frenkel E. P. et al. (1986) Therapeutic efficacy of multiagent chemotherapy with drug delivery enhancement of blood-brain-barrier modification in glioblastoma. Neurosurgery 19, 573-582. Schag C. C., Hemrick R. L. and Gruz P. A. (1984) Karnofsky performance status revisited: reliability. validity and guidelines. J. Clin. Oncol. 2, 187- 193.
0883s2897/91 $3.00+ 0.00 PergamonPressplc
1991
Printed in Great Britain.
Radionuclide Assessment of Blood-Brain Barrier Disruption Performed for Chemotherapy of High Grade Malignant Brain Gliomas AMOLAK
SINGH*,
BARBARA D. BELSHE and MARY K. GUMERLOCK
Departments of Radiology (Nuclear Medicine) and Neurosurgery, University of Missouri Hospital and Clinics, Columbia, MO 65212, U.S.A.
Compared to the conventional mode of chemotherapy of malignant brain gliomas following surgery and radiation therapy, chemotherapy after transient disruption of the blood-brain barrier coupled with intraarterial administration of methotrexate improved median survival from 12-14 to 22 months in our experience. Technetium-99m-DTPA brain scintigraphy played a unique and important role in the documentation of optimum blood-brain barrier disruption. Patients with excellent clinical outcome had significantly (P < 0.0005) better blood-brain barrier disruption than patients with poor outcome. The results indicate that the clinical outcome is related to the degree of blood-brain barrier disruption.
Introduction High grade malignant brain gliomas after initial surgical removal and radiation therapy are generally treated with chemotherapeutic agents which are administered orally or intravenously. Such therapeutic regimens have not significantly altered survival in these patients. At the University of Missouri Hospital and Clinics, the high grade malignant brain gliomas after initial surgery and irradiation are treated with chemotherapeutic drugs which are administered following transient disruption of the of the blood-brain barrier (BBB) using hyperosmolar intraarterial mannitol. By opening the blood-brain barrier at the time of chemotherapy the drug delivery to the tumor is enhanced not only by increasing concentration of the drug within the tumor mass but also by eliminating the steep concentration gradient which exists between tumor and surrounding brain tissue. As a result of this anti “SUMP” effect, the drug-tumor contact is prolonged. Moreover, to maximize tumor bombardment with chemotherapy, methotrexate is injected intraarterially directly into the blood supply of the tumor. The disruption of the blood-brain barrier is evaluated using 99mTc-diethylenetriaminepentaacetic acid (DTPA) brain scintigraphy to monitor this
*All correspondence should be addressed to: Amolak Singh, Division of Nuclear Medicine, One Hospital Drive, 2N-19 Columbia, MO 65212, U.S.A.
disruption tumor.
for optimum
drug chemotherapy
to the
Materials and Methods The study population included 37 patients with histologically proven high grade malignant brain gliomas. There were 15 female and 22 males, with a mean age of 43 years, ranging from 12 to 71 years. The functional condition of the patient was established by using Karnofsky’s performance (KP) score (Schag et al., 1984). A total of 246 chemotherapeutic treatments were instituted in 37 patients. (A) Technique for (BBBD)
blood-brain
barrier
disruption
BBBD was achieved using a hyperosmolar (25%) solution of mannitol injected into the internal carotid and/or vertebral arteries depending upon tumor blood supply. The rate of mannitol infusion varied from 5 to 12 mL/s for a 30 s period. Methotrexate (1-5 g) was infused intraarterially into the tumor blood supply either through the internal carotid artery or vertebral artery, or both in some cases, immediately following BBBD with hyperosmolar mannitol. Cyclophosphamide (20 mg/kg) was given intravenously 10 min prior to BBBD with mannitol to allow sufficient time for hepatic circulation and drug activation, Procarbazine (100 mg) was given orally daily for 14 days starting the day after disruption. 641
642
AMOLAK SINGH et al.
Leucovorin, a folic acid derivative, was used to counteract toxic effects of methotrexate. The patients underwent tumor BBBD-chemotherapy every 4 + weeks guided by clinical response, brain computed tomography (CT) and other laboratory studies done prior to each planned BBBD chemotherapy. (B) Technetium-9%
DTPA scintigraphy
All patients underwent a baseline brain scan and BBBD-brain scans during therapy. BBBD-brain scintigrams were obtained using 20mCi of 99mTcDTPA administered intravenously immediately following mannitol infusion. Anterior posterior, vertex and both lateral brain images were obtained 2-3 h post-injection of the radiopharmaceutical using a gamma camera (Siemens 7500) interfaced with a nuclear medicine computer (ADAC-3300). Occasionally, SPECT images were also generated in addition to routine planar images. Using the 99mTc-DTPA brain scintigram, the degree of BBBD was routinely evaluated using a grading system based on the presence or absence of increased brain uptake in the region outside the tumor margins. Absence of disruption was graded as 0, minimal uptake as I+, moderate uptake as 2 +, and 3 + as intense uptake equal to or greater than venous sinuses. The BBBD index, defined as the ratio of count density from the brain tissue in the vicinity but outside tumor margin to contralateral normal brain tissue, was determined. The BBBD index in patients with excellent outcome was compared with BBBD index in patients with poor outcome using Wilcoxon rank sum test. A P value co.05 was required for significant difference. All data were expressed as meanf 1 SD.
Results Using CT scan as a diagnostic test for evaluation of tumor size before therapy and subsequently during follow-up examinations, six (16%) patients were found to have excellent tumor response and were found alive 37-78 months following diagnosis. Twenty-four (65%) patients had partial or temporary remission. Seven (19%) had poor outcome. Calculated median survival time was 22 months. One hundred and forty-two (65%) out of 232 studies were judged to have adequate (2 + or 3 +) disruption, 54 (23%) had minimal (1 +) disruption and 27 (12%) had no disruption. BBBD was usually adequate with carotid artery injection of mannitol (72%) versus
vertebral artery injection in which case only 20% of the patients revealed adequate disruption. As shown in Table 1, patients with excellent response (Patient No. = 6, studies = 36) revealed a mean BBBD index of 1.94 f 0.25 compared to patients with poor outcome (Patient No. = 7, studies = 25) who exhibited a mean BBBD index of 1.42 + 0.25. The difference was highly significant (P c 0.0005). Representative cases (1) A 30-year-old female presented with a left frontal lobe high grade malignant glioma. A pretherapy CT scan revealed postoperative changes and residual tumor in the left frontal lobe [Fig. l(A)]. A BBBD scintigram demonstrated excellent (3 i-) BBBD at the time of chemotherapy [Fig. l(B)]. The patient underwent eleven such mannitol-induced BBBD and treatments. A follow-up CT scan showed complete resolution of the tumor [Fig. l(C)]. This patient was alive and well 4 years after the first treatment. (2) A 50-year-old female was included in the protocol for therapy of a large parietal lobe high glioma [Fig. 2(A)]. A BBBD brain scintigram demonstrated no significant BBBD outside the tumor margins [Fig. 2(B)]. The patient underwent three such mannitolinduced BBBD and chemotherapies without any response. A follow-up CT scan [Fig. 2(C)] exhibited further enlargement of the tumor. This patient died within a year following inclusion in the protocol.
Discussion Conventional mode of chemotherapy after surgical removal and irradiation therapy of high grade malig nant brain gliomas has not been successful in altering survival in these patients. Our results with BBBD chemotherapy indicate significant improvement in patient survival with increase in median survival time from 12-14 to 22 months. This compares favorably with 16.5 months reported by Neuwelt et al. (1986). Technetium-99m-DTPA brain scintigraphy plays an important role in the assessment of the degree of BBBD and documentation of optimum drug delivery to the area of interest (tumor) and the surrounding brain tissue. Moreover, scintigraphic studies were helpful for guiding mannitol infusion rates required for optimum BBBD. Adequate (2+ to 3+) BBBD occurred in 65% of our studies. No disruption was noted in 12%; in the remaining
Table I. Relationship between clinical outcome and degree of BBBD* Clinical outcome
BBBD Index (mean + SD)
Excellent response (Number of patients: 6; number of studies: 36)
1.94 f 0.44
Poor response (number of oatients: 7: number of studies: 25)
I .42 + 0.25
Pt
and of questionable significance. Internal carotid artery injections of mannitol produced adequate disruption in 72% of studies as compared to the vertebral artery injection in which case only 20% of the studies revealed adequate disruption. These results suggest that improved delivery of chemotherapeutic drugs to the region of interest as a result of BBBD is achieved in approximately three fourths (3/4) of treatments following mannitol injections through internal carotid artery, and in only one fifth (l/5) of treatments following vertebral artery injections of mannitol. However, all cases benefit from intraarterial administration of chemotherapeutic drugs. The degree of BBBD and clinical outcome appear related since patients with excellent outcome have better BBBD than those with poor outcome (P < 0.0005). This finding has not been previously reported. This application of radionuclide brain scan using 99mT~labeled chelates represents a unique application of conventional nondiffusible brain agents. Newer diffusible brain agents such as 99mTc-hexamethylpropyleneamine oxide (HMPAO) and ‘XI-labeled amphetamine derivatives are not suitable for this application. Moreover. the radionuclide brain scan is superior to CT in the assessment of BBBD because of higher safety. CT scans performed for the assessment
of BBBD are associated with fairly high (17%) incidence of contrast induced seizures in patients undergoing BBBD and chemotherapeutic treatments (Neuwelt et al., 1983). In conclusion, BBBD chemotherapy of malignant brain gliomas significantly improves survival compared to the conventional mode of chemotherapy. Technetium-99m DTPA (or other similar agents such as WmTc-glucoheptonate) brain scintigraphy is the technique of choice for the assessment of blood-brain barrier disruption, and for documentation of optimum drug delivery to the region of interest. The degree of BBBD as measured by brain scintigraphy appears related to survival and clinical outcome in these patients with high malignant brain glioma.
References Neuwelt E. A., Specht D. H., Howieson J. et al. (1983) Osmotic blood-brain-barrier modification: clinical documentation by enhanced CT scanning and/or radionuclide brain scanning. Am. J. Radiol. 4, 907-913. Neuwelt E. A., Howieson J., Frenkel E. P. et al. (1986) Therapeutic efficacy of multiagent chemotherapy with drug delivery enhancement of blood-brain-barrier modification in glioblastoma. Neurosurgery 19, 573-582. Schag C. C., Hemrick R. L. and Gruz P. A. (1984) Karnofsky performance status revisited: reliability. validity and guidelines. J. Clin. Oncol. 2, 187- 193.
